Cyclic Polycarbonates by Polycondensation of Bisphenol A with

Interfacial polycondensations of bisphenol A with triphosgene were conducted in the CH2Cl2/aqueous NaOH system. Four different catalysts were compared...
0 downloads 0 Views 156KB Size
1742

Macromolecules 2004, 37, 1742-1748

Cyclic Polycarbonates by Polycondensation of Bisphenol A with Triphosgene† Hans R. Kricheldorf,*,‡ Sigrid Bo1 hme,‡ Gert Schwarz,‡ and Claus-L. Schultz§ Institut fu¨ r Technische und Makromolekulare Chemie, Bundesstr. 45, D-20146 Hamburg, Germany, and Bayer AG, ZF-ZAU, Geba¨ ude R-79 (4. Etage), D-47812 Krefeld, Germany Received February 24, 2003; Revised Manuscript Received November 4, 2003

ABSTRACT: Interfacial polycondensations of bisphenol A with triphosgene were conducted in the CH2Cl2/aqueous NaOH system. Four different catalysts were compared: triethylamine, triethylbenzylammonium chloride (TEBA-Cl), tetrabutylammonium hydrogen sulfate, and tetraphenylphosphonium chloride. Triethylamine yielded mixtures of cyclic polycarbonates and OH-terminated polycarbonates. The fraction of cycles and the molecular weights increased with lower concentrations of this catalyst. When the pseudo-high-dilution method was applied, samples containing g95 mol % of cycles were obtained, and the average molecular weights were varied over a broad range via the feed ratio of triethylamine. Fractionation of one sample allowed the detection of cyclic polycarbonates up to 28 000 Da. The three phase transfer catalysts gave quite different product mixtures when compared under identical conditions. The highest fraction of cycles and the highest molecular weights were obtained with Ph4PCl. The fraction of cycles increased with higher feed ratios of Ph4PCl in contrast to triethylamine. Bimodal mass distributions were found for all samples with a high content of cyclic polycarbonates, whereas the frequency distributions were monomodal.

Introduction

Table 1. Interfacial Polycondensation of BPA with Triphosgenea Catalyzed with TEA

The high toxicity of phosgene combined with its gaseous character and a relatively pleasant smell in high dilution renders this versatile reagent a dangerous chemical. Two less dangerous substitutes have been commercialized over the past two decades, namely diphosgene (trichloromethyl chloroformate) and triphosgene (bis(trichloromethyl) carbonate). Whereas diphosgene contains a chloroformyl group like phosgene, triphosgene has a different structure, and thus, the question arises if the reactivity and preparative usefulness of triphosgene are comparable to those of phosgene. In numerous organic syntheses triphosgene has proven its usefulness.1 Furthermore, numerous polycarbonates were prepared by interfacial polycondensations with triphosgene.2-5 However, the influence of various catalysts on the course of these polycondensations has never systematically been studied. Another interesting aspect that has not been studied yet is the extent of cyclization in triphosgene-mediated syntheses of polycarbonates. In a recent study6 we have found that the results of interfacial polycondensations of bisphenol A by means of diphosgene are highly sensitive to slight variations of the reaction conditions. Careful optimization allows for the preparation of high molar mass polycarbonates with weight-average molecular weights (Mw) above 106 Da, unless a chain terminator is used. Furthermore, both molar fraction and ring size of cyclic oligo- and polycarbonates systematically increased with higher molecular weights of the entire sample. This tendency is characteristic for all kinetically controlled polycondensations as recently demonstrated for several different classes of polymers.7-9 In this context, it was one purpose of the present work to perform interfacial †

“Macrocycles”: Part 26. Institut fu¨r Technische und Makromolekulare Chemie. Bayer AG. * Corresponding author.

‡ §

expt no.

TEA/BPA

yield (%)

ηinhb (dL/g)

Meic (Da)

Mvd

1 2 3 4 5 6

0.5/21 1/21 2/21 4/21 8/21 12/21

79 72 61 67 50 41

2.50 2.85 2.60 0.67 0.25 0.19

>15 000 >15 000 >15 000 3300 2000 1600

130 000 140 000 135 000 25 000 13 000 12 000

a Initial temperature +5 °C, molar NaOH/BPA ratio 126/21 molar triphosgene/BPA ratio 14/21. b Measured at 20 °C with c ) 2 g/L in CH2Cl2. c Mass where cyclic and linear polycarbonates show equal peak intensities in the MALDI-TOF MS. d Calculated from the ηinh values via the MH equation (14).

polycondensations of bisphenol A with triphosgene and to study the influence of various catalysts on the efficiency of the chain growth and on the extent of cyclization. Another purpose was the synthesis of cyclic polycarbonates of low or medium molecular weight useful for processing from the melt. Experimental Section Materials. Bisphenol A was a gift of Bayer AG (Leverkusen, Germany) and was used after recrystallization (purity >99% by GC). Triphosgene was purchased from Lancaster Synthesis (Clariant, Frankfurt, Germany) and was used as received. CH2Cl2 was also a gift of Bayer AG and was distilled over P4O10 before use. Triethylamine was a gift of BASF AG (Ludwigshafen, Germany) and was used as received. Benzyltriethylammonium chloride (TEBA-Cl), tetrabutylammonium hydrogen sulfate, and tetraphenylphosphonium chloride were all purchased from Aldrich Co. (Milwaukee, WI) and used as received. Polycondensations. (A) With Triethylamine (Table 1). Bisphenol A (21 mmol) and NaOH (126 mmol) were dissolved in H2O (200 mL) and cooled to +5 °C in a refrigerator. Triphosgene (14 mmol) was dissolved in dry CH2Cl2 (200 mL) and cooled to +5 °C. Immediately before both solutions were mixed with a high-speed stirrer (“Ultraturrax”), the triethylamine (2 mmol) was added to the water phase. During the high-speed

10.1021/ma0301348 CCC: $27.50 © 2004 American Chemical Society Published on Web 02/11/2004

Macromolecules, Vol. 37, No. 5, 2004

Cyclic Polycarbonates 1743

Table 2. Triethylamine-Catalyzed Interfacial Polycondensationsa of Bisphenol A with Triphosgene via the Pseudo-High-Dilution Method at 20 °C expt no.

TEA/Bisphenol A

yield (%)

ηinhb (dL/g)

ηintrc (dL/g)

Mnd,e (Da)

Mwd (Da)

Tgf (°C)

1 2 3 4 5

0.25/21 0.50/21 1.00/21 2.00/21 4.00/21

73 72 71 70 69

1.03 1.51 0.45 0.36 0.34

1.08 1.53 0.48 0.37 0.35

10 000 14 500 4 500 3 500 3 000

95 000 175 000 36 000 28 000 26 000

148.5 150.5 146.0 146.0 139.5

a Initial reaction temperature +5 °C, molar NaOH/BPA ratio 126/21, molar triphosgene/BPA ratio 14/21. b Measured at 20 °C with c ) 2 g/L in CH2Cl2. c Measured at 30 °C in CH2Cl2. d SEC measurements in CH2Cl2 evaluated with a triple detector. e Because of the broad bimodal MWD the margin of error is on the order of (15-20%. f DSC measurements with a heating rate of 20 °C/min (first heating).

Table 3. Characterization of the Ten Fractions Obtained by SEC from Sample No. 3, Table 2 SEC measurements no. 1 2 3 4 5 6 7 8 9 10

[η]a

(dL/g)

2.143 1.508 0.943 0.557 0.327 0.243 0.102

Mwb 260 000 135 000 66 000 33 500 21 000 12 500 8 100

Mpc 243 000 122 000 62 000 32 000 19 500 11 200 6 200

MALDI-TOF Mpd

29 000 14 300 7 400 3 300 1 560 1 050 8000

a Initial temperature +5 °C, molar NaOH/BPA ratio 126/21, molar triphosgene/BPA ratio 14/21. b Measured at 20°C with c ) 2 g/L in CH2Cl2. c Mass where cyclic and linear polycarbonates display equal peak intensities in the MALDI-TOF MS. d Mn ) 11 100 Da, Mw ) 33 000 Da by SEC with triple detection. e Mn ) 7000 Da, Mw ) 29 000 Da by SEC with triple detection. f Mn ) 17 000 Da, Mw ) 55 000 Da by SEC with triple detection.

Figure 1. MALDI-TOF mass spectrum of the polycarbonate no. 4, Table 1. principle agreed with those obtained from products of the standard workup procedure, but strong peaks of Na+-doped products were present in addition to the K+-doped peaks with the consequence of lower signal-to-noise ratios. Measurements. The inherent viscosities were measured in CH2Cl2 with an Ubbelohde viscometer thermostated at 20 °C. The MALDI-TOF mass spectra were recorded with a Bruker Biflex III equipped with a nitrogen laser (λ ) 337 nm) in the reflectron mode. An acceleration voltage of 20 kV and cutoff range of 1000 or 4000 Da were used. The irradiation targets were prepared from CH2Cl2 solutions with dithranol as matrix and K-trifluoroacetate as dopant. The SEC were performed on a Hewlett-Packard HP 1050 apparatus in CH2Cl2 at 30 °C. Five Lichrogel columns having pore sizes of 4, 40, 400 (2×), and 4000 Å were used. The elution curves were evaluated with a triple detector “Viskotek-TDA 301” combined with the software “Viskotek Tri SEC”.

1744 Kricheldorf et al.

Figure 2. MALDI-TOF mass spectrum of the polycarbonate no. 3, Table 2, prior to its fractionation.

Results and Discussion Triethylamine (TEA) as Catalyst. In a first series of polycondensations TEA was used as catalyst because this amine had proven to be particularly useful in syntheses of cyclic oligo- and polycarbonates via hydrolytic polycondensation of bisphenol A bis(chloroformate).10-13 Furthermore, TEA has also proven its usefulness in previous studies of interfacial phosgenations of bisphenol A with diphosgene.5 The reaction conditions were selected to match those used in a previous study based on phosgenation with diphosgene. The molar ratio of triphosgene/BPA corresponded to a 100% molar excess of COCl2, and the NaOH/BPA ratio was designed for complete neutralization of the liberated HCl and 50% neutralization of H2CO3. Since an initial temperature below 20 °C has proven to favor propagation and cyclization over hydrolysis of (di)phosgene,6 the initial temperature used in this work for the first series of experiments (Table 1) was again fixed at +5 °C. Whereas all these parameters were kept constant, the TEA/BPA ratio was varied. The results listed in Table 1 demonstrate that the inherent viscosities pass through a maximum at a low TEA feed ratio. An analogous trend was found for phosgenations with diphosgene.6 Even the yields follow this trend, although the workup procedure was not optimized for high yield but for Na-free polycarbonates. Na-free polymers were desirable to obtain optimum MALDI-TOF mass spectra (MS) with K+ doping. The MALDI-TOF MS revealed the existence of cyclic polycarbonates in all samples (nos. 1-6, Table 1). When the molecular weight was high, only mass peaks of

Macromolecules, Vol. 37, No. 5, 2004

Figure 3. MALDI-TOF mass spectrum of fraction no. 5 of polycarbonate no. 3, Table 2.

cycles were detectable (nos. 1-3). Their fraction decreased with lower molecular weights, and the linear chains of structure La (Scheme 1) became the predominant reaction products. This trend is illustrated in Figure 1 and by the Mei values listed in Table 1. Mei means the mass value where the mass peaks of cycles and linear chains have equal intensities. Lower Mei values mean lower content of cyclic oligo- and polycarbonates. The finding that higher molecular weights are combined with higher fractions of cycles is in perfect agreement with our recent theory of step-growth polymerization.7-9 These results also provide a clue to the polymerization mechanism (together with the results obtained from phase transfer catalysts below). Considering the well-known14-17 complexation of chloroformate groups with triethylamine, it may be assumed that triethylamine reacts with triphosgene according to eq 1 (see Scheme 2). The acylammonium ion 1 is hydrophilic; it will diffuse into the water phase and hydrolyze (eq 2). Alternatively, it will react with phenoxide ions or with phenol groups (eq 3). The resulting mixed carbonate can react with a phenoxide ion (eq 4) to yield a stable carbonate group (eq 5). Another reaction sequence will start out from the decomposition of the trichloromethoxide ion into chloride ions and phosgene (eq 6). The phosgene will be activated by triethylamine (eq 7), and the acylammonium ion 2 which is more hydrophilic than 1 has in turn the choice between hydrolysis and formation of a phenyl chloroformate group (eq 8). This chloroformate group can react with a phenoxide ion yielding a stable carbonate (eq 9). On the other hand, an activation of the chloroformate with TEA

Macromolecules, Vol. 37, No. 5, 2004

Cyclic Polycarbonates 1745 Scheme 2

Figure 4. MALDI-TOF mass spectrum of the polycarbonate no. 1, Table 4 (A), no. 2, Table 4 (B), and no. 5, Table 4 (C). Scheme 1

may take place (eq 10), and the acylammonium ion 3 has then the choice between hydrolysis and continuation of the chain growth (eq 11). Even though eqs 1-11 are not an exhaustive description of all possible reaction steps, they illustrate the role of TEA. The experimental results indicate that higher feed ratios of TEA favor the hydrolysis of triphosgene (and its reaction products) via hydrophilic acylammonium ions more than chain growth and cyclization.

TEA and the Pseudo-High-Dilution Method. The results discussed above and those obtained with diphosgene6 have demonstrated that polycarbonates with a high percentage of cycles can be prepared when the reaction conditions favor high molecular weights (weightaverage molecular weights, Mws, in the range 105-106 Da). However, polycarbonates having Mws above 50 000 Da possess too high a melt viscosity for processing. Therefore, it was of interest to find reaction conditions allowing for the synthesis of low molar mass cyclic polycarbonates. For this purpose, a second series of polycondensations were conducted using the pseudohigh-dilution method in combination with a higher reaction temperature. The feed ratio of TEA was varied, whereas all other parameters were kept constant. The results summarized in Table 2 allow the following conclusions. First, the molecular weights pass through a maximum around 2.5 mol % of TEA (relative to BPA). Compared to the polycondensations performed at higher concentrations (Table 1), the maximum of the molecular weights is shifted to a lower feed of TEA. Second, all molecular weights are shifted to lower values as expected for the pseudo-high-dilution method. However, at the optimum TEA feed, the molecular weight of the isolated polycarbonate (no. 2, Table 2) is still too high

1746 Kricheldorf et al.

for a convenient processing. Third, the most interesting and important result is the observation that the MALDITOF MS of all samples are nearly identical, despite a broad variation of the molecular weights. Even in the case of sample no. 4 (Table 2) only traces of linear chains (La type) are detectable, whereas no linear chains were found at masses above 3000 Da. This trend is quite different from that found for the samples of Table 1 as illustrated by a comparison of the MS presented in Figures 1 and 2. This comparison demonstrates that the pseudo-high-dilution method favors the cyclization even at low molecular weights. The MS presented in Figure 2 demonstrates that the signal-to-noise ratio is poor above 10 000 Da. To obtain more information about the structure of the polycarbonate chains above 10 000 Da, sample no. 3, Table 2, was fractionated by SEC, so that 10 fractions were obtained. The characterization of these fractions is summarized in Table 3. These fractions were characterized by MALDI-TOF mass spectrometry, and from five fractions satisfactory mass spectra were obtained, whereas in the case of fraction 4 the signal-to-noise ratio was poor. In all MS, only mass peaks of cyclic oligo- and polycarbonates were observable. The best mass spectrum is displayed in Figure 3. Cyclic polycarbonates were detectable up to 28 000 Da. Considering the relatively low Mn and Mw values of this sample, it may be concluded that certainly more than 90 mol % of this sample consists of cycles. (The weight percentage is necessarily lower than the mole percentage because the longest chains have the lowest tendency to cyclize.) This estimation is based on the following consideration: The majority of cycles (around 65-70 mol %) is concentrated in the mass range of the frequency distribution below Mn. Around 90 mol % are concentrated in the mass range of the frequency distribution below Mw. Furthermore, the Mn and Mw values obtained by triple detection overestimate the true molecular weights in the mass range below 30 000 Da, as evidenced by a comparison of MALDI-TOF and SEC measurements of the fractions listed in Table 3. The reason for this discrepancy is the fact that the absolute molecular weights of the triple detection are based on light scattering, which is insensitive to the presence of (cyclic) oligomers and low molar mass polymers. For the other samples listed in Table 2, a similarly high content of cyclic oligo- and polycarbonates may be assumed as for no. 3. Therefore, these results prove that triphosgene is particularly well suited to prepare cyclic polycarbonates with variable molecular weights by direct polycondensation of bisphenol A. Considering that the content of cycles is quite similar in all four samples of Table 2, these samples are well suited to study the dependence of chemical and physical properties on the molecular weight. In this work, only glass transition temperatures (Tgs) should be reported which play a key role for the heat distortion temperature and thus for commercial applications of the polycarbonate. The Tgs listed in Table 2 display the expected dependence on the molecular weight, but a significant deviation from the Tgs of commercial linear polycarbonates having similar molecular weights was not detectable. Phase Transfer Catalysis. The successful use of phase-transfer (PT) catalysts in the interfacial phosgenation of various diphenols by means of monomeric phosgene is well documented in the literature.18,19 In a

Macromolecules, Vol. 37, No. 5, 2004

Figure 5. SEC elution curves of the polycarbonates no. 4, Table 2 (A), no. 3, Table 2 (B), and no. 1, Table 2 (C). Scheme 3

previous study6 based on diphosgene triethylbenzylammonium chloride (TEBA-Cl) and tetrabutylammonium (TBA) hydrogen sulfate were used. Both PT catalysts gave similar results. The content of cyclic polycarbonates increased with higher molecular weights of the reaction products. In the present work, these PT catalysts were again used, and tetraphenylphosphonium (TPP) chloride was also included in this study. All three PT catalysts were compared under identical conditions (nos. 1, 2, 5; Table 4), and these reaction conditions corresponded to the standard conditions also used for the experiments of Table 1. Surprisingly, three different reaction products were obtained as evidenced by the MALDI-TOF MS presented in Figure 4. With TEBA-Cl only few cycles were formed, and the vast majority of the products consisted of linear chains having two OH end groups (La). In the original MS, the mass peaks of these La chains were detectable up to 18 000 Da. In the case of TBA hydrogen sulfate, a variety of different products were formed which were

Macromolecules, Vol. 37, No. 5, 2004

Cyclic Polycarbonates 1747

Molecular Weight Distributions (MWDs). In our previous work on polycarbonates,6 a Mark-HouwinkSakurada equation

[η] ) 0.036Mw0.70

Figure 6. SEC elution curves of the polycarbonates no. 1, Table 4 (A), and no. 5, Table 4 (B).

not analyzed in detail. With TPP-Cl mainly cyclic oligoand polycarbonates were formed. When compared to the TEA-catalyzed sample no. 5 of Table 1, the content of cycles was even higher in the case of the TPP-Cl catalysis. Particularly interesting proved the trend observed for the variation of the TPP-Cl/BPA ratio. A higher feed of TPP-Cl gave higher molecular weights in combination with a higher content of cycles. This trend is opposite to that found for TEA (Tables 1 and 2), indicating another polymerization mechanism. It is obvious that TPP-Cl (like other PT catalysts) cannot activate triphosgene, phosgene, or chloroformate groups in a manner analogous to TEA (eqs 1, 7, and 10). Hence, the catalytic effect of TPP-Cl must be due to an activation of the phenoxide ion. Besides fluoride ions, oxide ions are the strongest acceptors of H-bonds and, therefore, surrounded by a rather stable shell of H2O molecules when dissolved in water. The transfer to the organic phaseremoves most or all of the water molecules, as schematically outlined in eq 12 (see Scheme 3). This change of the solvation shell tremendously enhances the nucleophilicity, and this change is the basis of phasetransfer catalysis.20,21 The TPP cation will also transfer more or less hydrated OH- ions to the organic phase,21 thereby catalyzing the hydrolysis of triphosgene, phosgene, and chloroformate groups. However, the phenolic OH groups are about 6 orders of magnitude more acidic than water, so that the concentration of phenoxide groups in the organic phase should be much higher than that of hydroxide ions. In consequence, increasing feed ratios of TPP-Cl favor propagation (eq 13) and cyclization over hydrolysis of the electrophilic reaction partners. A similar trend was observed for TEBA-Cl when diphosgene was polycondensed with bisphenol A.6 In summary, the results obtained in this work demonstrate that the structure of the PT catalyst has an enormous influence on the composition of the reaction product. Furthermore, the reactivity of triphosgene toward PT catalysts is different from that of diphosgene and thus justifies a separate investigation.

(14)

was elaborated for polycarbonates of bisphenol A mainly consisting of cycles. This equation was used to calculate the viscosity-average molecular weights listed in Table 1 which had the purpose to illustrate the order of magnitude. These Mv values were calculated from the inherent viscosities which usually deviated from intrinsic viscosities by 5% as found by comparison of numerous viscosity measurements.6,13,22 A higher accuracy of the Mv calculations is not more informative because the MWDs vary with the reaction conditions as demonstrated below and in a previous publication.6 Samples nos. 1-5 of Table 2 and nos. 1 plus 5 of Table 4 were subjected to SEC measurements which were evaluated by the triple-detection method including light scattering. The samples of Table 2 displayed a bimodal mass distribution regardless of their molecular weight (Figure 5). Similar bimodal mass distribution were found in previous studies13 of hydrolytic polycondensation of bisphenol A carbonate, whenever high fractions of cycles were formed. Similar bimodal mass distributions were also observed for syntheses of poly(ether sulfone)s under conditions favoring the formation of cycles.23 Hence, it may be concluded that the bimodal mass distribution is not a consequence of the interfacial polycondensations. This conclusion is supported by the results obtained with PT catalyst. With TEBA-Cl (no. 1, Table 4) only few cycles were formed (Figure 4A), and the SEC curve (Figure 6A) displays a monomodal character. Under the same conditions Ph4PCl yielded a much higher fraction of cycles (Figure 4C), and the SEC curve displayed a “bimodal trend” (Figure 6B). These bimodal MWDs have, of course, the consequence of high polydispersities, and the margin of error of the numberaverage molecular weights (Mns) is particularly large. The relationship between permanent cyclization and bimodal mass distribution can be rationalized considering that the cylic oligomers mainly formed in an early stage of a polycondensation are stable under the conditions of a KCP and do not vanish. In contrast, Flory’s original concept of step-growth polymerizations24 does not include a significant contribution of cyclization reactions and predicts a monomodal mass distribution according to the equation

Wn ) mpn(1 - p)

(15)

where Wn ) (weight fraction of oligomers)/(polymers having a degree of polymerization n), m ) molar mass of the repeating unit, and p ) conversion. This concept is based on the assumption that all (linear) oligomers gradually vanish with increasing conversion. The results presented here and in other recent publications clearly prove that kinetically controlled polycondensations do not obey Flory’s original treatment of MWDs. However, it should also be mentioned that Flory calculated later25 a bimodal mass distribution for thermodynamically controlled polycondensations involving a formation of cyclic oligomer via “backbiting” equilibration. This later modification of Flory’s theory may be considered as an indirect confirmation of our hypothesis that when

1748 Kricheldorf et al.

intensive cyclization occurs in a KCP a bimodal mass distributions will result. Conclusion The results elaborated in this work allow interesting conclusions in three directions. First, the reactivity of triphosgene is so different from that of diphosgene that at least with PT catalysts interfacial polycondensations of diphenols give results which are quite different from those obtained with diphosgene under identical conditions.5,21 Second, triphosgene allows one to prepare cyclic oligo- and polycarbonates of bisphenol A in a simple “one-pot-procedure” based on the pseudo-highdilution method and triethylamine as catalyst. Variation of the catalyst feed and of the temperature (documented in more detail in a future publication) allows one to vary the average molecular weight of the cyclic polycarbonates over a broad range. Third, the SEC elution curves prove together with those recently obtained from other polymer syntheses12,21,22 that reaction conditions yielding a high percentage of cycles automatically produce a bimodal mass distribution along with a monomodal frequency distribution. These MWDs differ from those calculated by Flory25 in his classical theory for polycondensations, where a significant influence of cyclization reactions on kinetically controlled polycondensations was not taken into account. Further studies designed to elucidate both the experimental and the theoretical background of MWDs of cyclic polymers are in progress. References and Notes (1) Eckert, H.; Forster, B. Angew. Chem. 1987, 99, 922; Angew. Chem., Int. Ed. Engl. 26, 894. (2) Kricheldorf, H. R.; Lu¨bbers, D. Makromol. Chem. Rapid Commun. 1989, 10, 383. (3) Kricheldorf, H. R.; Lu¨bbers, D. Macromolecules 1990, 23, 2656. (4) Sun, S. J.; Hsu, K.-Y.; Chang, T.-C. Polym. J. 1997, 29, 25.

Macromolecules, Vol. 37, No. 5, 2004 (5) Marks, M. J.; Munja, S.; Namhatu, S.; Scott, D. C.; Bosscher, S.; De Letter, J. A.; Klumperman, B. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 560. (6) Kricheldorf, H. R.; Bo¨hme, S.; Schwarz, G.; Schultz, C.-L. J. Polym. Sci., Part A: Polym. Chem., in press. (7) Kricheldorf, H. R.; Rabenstein, M.; Maskos, M.; Schmidt, M. Macromolecules 2001, 34, 713. (8) Kricheldorf, H. R.; Bo¨hme, S.; Schwarz, G. Macromolecules 2001, 34, 8879. (9) Kricheldorf, H. R.; Bo¨hme, S.; Schwarz, G.; Kru¨ger, R.-P.; Schulz, G. Macromolecules 2001, 34, 8886. (10) Horbach, A.; Vernaleken, H.; Weirauch, K. Makromol. Chem. 1980. (11) Brunelle, D. J.; Boden, E. P.; Schannon, T. G. J. Am. Chem. Soc. 1990, 112, 2399. (12) Brunelle, D. J.; Schannon, T. G. Macromolecules 1991, 24, 3035. (13) Kricheldorf, H. R.; Bo¨hme, S.; Schwarz, G.; Schultz, C.-L. Macromol. Rapid Commun. (14) Schnell, H. Angew. Chem. 1956, 68, 633. (15) Vernaleken, H. Polycarbonates. In Interfacial Synthesis; Millich, F., Carrakeer, C. E., Jr., Eds.; Marcel Dekker: New York, 1977; Vol. 2. (16) Kosky, P. G. J. Polym. Sci., Part A: Polym. Chem. 1991, 29, 101. (17) Aquino, E. C.; Brittain, W. J.; Brunelle, D. J. J. Polym. Sci., Part A: Polym. Chem. 1994, 32, 741. (18) Tagle, L. H. In Handbook of Phase Transfer Catalyses; Sasson, Y., Neumann, R., Eds.; Blackie Academic & Professional: London, 1997; Chapter 6. (19) Ide, K.; Serini, V.; Freitag, D.; Tengler, G. Ger. Offen. 2901 665 (1980) to Bayer AG. (20) Dehmlow, E. V.; Dehmlow, S. S. Phase Transfer Catalysis, 3rd ed.; Verlag Chemie: Weinheim, 1993. (21) Esikova, T. A. In Handbook of Phase-Transfer Catalysis; Sasson, Y., Neumann, R., Eds.; Blackie Academic & Professional: London, 1997; Chapter 1. (22) Kricheldorf, H. R.; Bo¨hme, S.; Schwarz, G.; Schultz, C.-L., manuscript in preparation. (23) Kricheldorf, H. R.; Vakhtangishvili, L.; Schwarz, G.; Kru¨ger, R.-P. Polymer, in press. (24) Flory, P. J. Chem. Rev. 1946, 39, 137. (25) Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953; Chapter VIII.

MA0301348